Measurement Machine With Three Zeiss Coordinates: How It Detects CNC Hardware Parts

• 1
  What a measurement machine with three zeiss coordinates actually measures
• 2
  Part alignment decides whether the numbers mean anything
• 3
  Measurement uncertainty and the tolerance budget
• 4
  Where a three-axis CMM falls short
• 5
  Which inspection method fits which part

A measurement machine with three zeiss coordinates is a bridge or gantry CMM. Three linear axes move a Zeiss probe in X, Y and Z. Each axis carries a glass scale, and the controller reads all three positions at the exact moment the probe tip contacts the surface. The result is a single point in 3D space, accurate to microns.

One point tells you almost nothing. The value comes from hundreds or thousands of points. Touch 8 points on a bore and the software fits a circle or cylinder. Touch 12 points on a plane and it fits a plane. From those fits come diameter, roundness, perpendicularity, position and flatness.

That is the core difference from a caliper or micrometer. Hand tools measure a single dimension between two contact faces. A CMM measures the geometry of the feature and compares it to the CAD model or drawing. It answers whether the hole is in the right place, not just whether it is the right size.

The probe is the other half of the system. Zeiss scanning probes glide along the surface and collect points continuously, while touch-trigger probes record discrete hits. Scanning gives more data on form; touch-trigger is faster for simple position checks. Both run on the same three-axis frame.

• 1
  Point cloud, not one dimensionHundreds of points build a feature fit, so form errors show up.
• 2
  CAD comparisonMeasured geometry is overlaid on the model to show deviation.
• 3
  Feature relationshipsPosition, perpendicularity and concentricity come from fitted features.

Before any dimension is reported, the machine needs to know where the part sits. This is the alignment, and it is the step that separates useful CMM data from noise. A three-axis machine does not find the part on its own. The operator defines a coordinate system from datums on the part.

A typical alignment uses a primary datum plane, a secondary datum axis and a tertiary datum point. For a machined housing, that might be the mounting face, a dowel bore and a locating slot. The software levels to the plane, rotates to the bore and translates to the slot. Only then does it report position.

If the alignment uses the wrong datum, every dimension shifts. A hole that is truly in position can read 0.08 mm out simply because the part was leveled to a rough cast surface instead of the finish-machined face. We see this most often on first-article parts where the drawing datum scheme is not obvious.

For production runs, we build a fixture that repeats the same datum contact every time. A dedicated fixture cuts setup time and removes operator variation. On high-volume hardware, the fixture can pay for itself in the first few hundred parts.

• 1
  Primary datumUsually a machined face that stops three degrees of freedom.
• 2
  Secondary datumA bore or edge that stops two more.
• 3
  Tertiary datumA slot or point that locks the last one.

Every CMM has a stated uncertainty, often written as MPE or maximum permissible error. A typical bridge machine in a temperature-controlled room holds around 1.5 + L/350 μm, where L is the measured length in millimeters. At 100 mm, that is roughly 1.8 μm. At 1,000 mm, it grows to about 4.4 μm.

This matters because the CMM error eats into the part tolerance. If a bore is specified at ±0.010 mm and the machine uncertainty is 2 μm, the measurement only leaves 8 μm of real margin. The rule of thumb is a 4:1 or 10:1 ratio between tolerance and measurement uncertainty. Below 4:1, you are arguing about noise.

Temperature is the largest practical error source. Steel grows about 11 μm per meter per °C. Aluminium grows about 23 μm. A 5 °C swing between the machine and the part can shift a 200 mm aluminium dimension by more than 20 μm. That is larger than most tight tolerances on the drawing.

For work in the ±0.005 mm range, we let parts soak in the inspection room until they reach 20 °C. The soak time depends on mass. A small aluminium bracket may need 30 minutes; a heavy steel block can need several hours.

• 1
  Soak before measuringParts and machine at the same temperature, usually 20 °C.
• 2
  Match the ratioKeep measurement uncertainty at least 4× smaller than tolerance.
• 3
  Log the environmentRecord temperature and humidity with the inspection report.

A measurement machine with three zeiss coordinates is a great general-purpose tool, but it is not universal. It cannot see inside a deep, narrow pocket that the stylus cannot reach. It cannot measure a feature on the underside of an overhang without repositioning the part. Every re-fixturing adds setup time and a new source of error.

It also cannot measure soft or flexible parts in a meaningful way. Touch force deflects rubber, thin-wall plastic and some thin sheet metal. The probe reads the deflected shape, not the free shape. For those parts, optical or non-contact scanning is the better choice.

Speed is the other limit. A three-axis machine moves one point at a time. A complex aerospace bracket with 300 features can take hours per part. That is fine for first article and sampling, but it does not fit a 10,000-piece production line without a sampling plan.

Finally, a CMM only reports what it is programmed to check. If the inspection program misses a feature or uses the wrong nominal, the report looks clean while the part is wrong. That is a programming failure, not a machine failure, and it is the most common reason a bad part ships.

• 1
  Reach limitsDeep pockets and undercuts may need a different probe or setup.
• 2
  Soft materialsTouch force deforms rubber, thin plastic and foil.
• 3
  Cycle timeComplex parts take hours; use sampling for high volume.
• 4
  Program coverageA missed feature is invisible in the report.